Recent data demonstrate that unrelieved cancer pain is a significant problem in the ambulatory oncology population. Patients report experiencing moderate to severe pain that interferes with functional abilities and has deleterious effects on quality of life. Our data, as well as work by others, suggest that research must be done to test the effectiveness of specific interventions to improve the outpatient's ability to more effectively manage cancer-related pain. Therefore, this study will test the effectiveness of an intervention designed using self- care theory, as well as the principles of "academic detailing" and nurse coaching, compared to the provision of only written information about pain management using the consumer version of the AHCPR cancer pain guideline, in decreasing the patients' pain intensity score, improving the patients' functional status, and improving the patients' quality of life. Adult oncology patients (N=320) experiencing pain from bone metastasis will be recruited from 8 outpatient sites in urban and rural areas of California. Patients will be stratified by site and based on whether or not they have an identified family caregiver, who agrees to participate in the study. Patients without a family caregiver will be randomized to either the PRO- SELF(C): PAIN CONTROL group or the AHCPR group. Patients with a family caregiver will be randomized to either the PRO-SELF(C):PAIN CONTROL group or the AHCPR group. The patient and the family caregiver will receive the intervention together. Baseline data will be collected on all patients and family caregivers. Patients and patient-caregiver dyads, in the PRO-SELF(C) groups, will receive the PRO-SELF(C):PAIN CONTROL intervention, for 6 weeks, during office visits and follow-up coaching phone calls. Weekly data collection will include measurement of pain intensity. Patients and patient-caregiver dyads, in the AHCPR groups, will be called and seen with the same frequency as the PRO-SELF(C) groups. Both of the PRO-SELF(C) groups and the AHCPR groups will be seen in the office at the end of week 6 to complete the data collection. Data will be analyzed using parametric statistics to determine if there is a decrease in the patients' pain intensity score, an improvement in functional status, and an improvement in quality of life in both PRO-SELF(C) groups compared to the AHCPR groups (i.e., patient or patient/caregiver dyad).